<![CDATA[The issue of divorce remains a complex and sensitive topic within Indonesian society, influenced by various factors such as repeated disputes, domestic violence, lack of harmony, financial difficulties, and other socio-cultural aspects. With the rise of social media, particularly Twitter, public discussions regarding divorce have become more widespread, allowing individuals to express their opinions and sentiments on the subject. These diverse perspectives create a wealth of sentiment data that can be analyzed to understand public perception and societal trends related to divorce. This study aims to classify public sentiment on divorce-related discussions using the Random Forest algorithm, providing insight into how people perceive and react to divorce issues. The research adopts a quantitative approach with a case study framework. The methodology involves data collection through web scraping techniques to gather approximately 1500 tweets containing discussions on divorce. The collected data is then preprocessed, including text cleaning, tokenization, and feature extraction, before being used to train and evaluate the Random Forest model. Sentiments are classified into three categories: negative, neutral, and positive. The classification model's performance is assessed using accuracy and F1-score metrics derived from the confusion matrix to determine its effectiveness in categorizing sentiments. Experimental results indicate that the Random Forest algorithm achieves an accuracy of 70%. The relatively low accuracy is attributed to the imbalance in sentiment class distribution, where negative sentiments dominate while positive sentiments are underrepresented. This imbalance affects the model's ability to predict positive sentiments effectively. The implications of this research contribute to a better understanding of public sentiment dynamics regarding divorce, which can be beneficial for policymakers, psychologists, and social researchers in analyzing societal attitudes towards marital dissolution.]]>
